Effect of CaCO3 Aqueous Saturation on CO2 Corrosion of ...

Investigations on the CO2 Corrosion of Mild Steel in the Presence of Magnesium and Calcium Ions

Hamed Mansoori, David Young, Bruce Brown, Srdjan Nesic, Marc Singer Institute for Corrosion and Multiphase Technology,

Department of Chemical & Biomolecular Engineering, Ohio University 342 West State Street

Athens, OH 45701 USA

ABSTRACT

Oilfield brine contains magnesium (Mg2+) and calcium (Ca2+) ions. Although carbon dioxide (CO2) corrosion of mild steel and its corrosion products has been extensively studied, the effect of Mg2+ and Ca2+ are frequently overlooked by researchers. Usually, a simple test electrolyte with various percentages of dissolved sodium chloride (NaCl) has been employed to study the mechanism of CO2 corrosion to avoiding complications relating to the water chemistry. The current paper is a continuation of a series of publications aimed at investigating the effect of water chemistries associated with Mg2+ and Ca2+ on the CO2 corrosion mechanism. A combination of electrochemical methods, weight loss technique, and surface characterization tools were employed over the course of long-term experiments with a controlled water chemistry and well-defined mass transfer conditions. The presence of Ca2+ in the electrolyte and the formation of a uniform calcium carbonate (CaCO3) layer on the steel surface facilitated precipitation of a protective iron carbonate (FeCO3) adjacent to the substrate. Localized corrosion was observed in the presence of high concentration of Mg2+ but was not detected in the simultaneous presence of Ca2+ and Mg2+ with a typical concentration ratio observed in oilfields (Ca2+/Mg2+=7).

Key words: CO2 corrosion, Mild steel, Magnesium, Calcium, Corrosion Product

1

Paper No.C2020-14914

©2020 by NACE International.Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing toNACE International, Publications Division, 15835 Park Ten Place, Houston, Texas 77084.The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

2

INTRODUCTION

Brine coproduced with a hydrocarbon phase usually contains high concentrations of magnesium ions (Mg2+) and calcium ions (Ca2+). Indeed, such brines are typically saturated with respect to calcium carbonate (CaCO3) and/or magnesium carbonate (MgCO3), with varying concentrations depending on environmental and geologic conditions. Precipitation of carbonate scales will occur due to changes in operational parameters (e.g., pH, pCO2, and temperature) when the fluid is transferred from the reservoir to surface facilities. It is worth mentioning that a pure MgCO3 scale has rarely been observed in oilfields. Instead, when brine contains Mg2+ and Ca2+, precipitation of a substitutional solid solution of calcium-magnesium carbonate (CaxMg1-xCO3) has been reported. This could be due to the fact that MgCO3 precipitation kinetics is extremely slow compared to CaCO3.1–3

Therefore, when the system is thermodynamically supersaturated with respect to both of these carbonates, CaCO3 nucleation and precipitation occur earlier. However, Mg2+ can incorporate into the lattice of CaCO3 during the crystallization process because MgCO3 (magnesite) and CaCO3 (with which it shares a calcite-type structure) are isostructural. Mechanisms of CO2 corrosion of mild steel and the characteristics of its corrosion products (mainly FeCO3 and Fe3C) have been extensively studied, and documented.4–8 However, most of these studies have been performed in various dilute solutions of sodium chloride (NaCl) while, in reality, Mg2+ and Ca2+ are also present in the brines produced from oil fields.9 A recent review paper by the current authors has investigated the relevant literature and the existing gaps in understanding the effect of Mg2+ and Ca2+ on mild steel corrosion.10 Mg2+ and Ca2+ are considered non-electroactive species, therefore, they do not directly participate in the electrochemical reactions involved in corrosion processes. However, they can alter the physiochemical properties of corrosion products and thus influence the corrosion mechanism. In a series of publications by the authors, the effect of Ca2+ and CaCO3 scale on aqueous CO2 corrosion of mild steel was studied for different scenarios in controlled water chemistry and well-defined mass transfer conditions.11–13 The following conclusions can be drawn from the conducted experiments at different CaCO3 saturation levels, [Ca2+], ionic strength, and pH at

80C, pCO2 0.53 bar, in 1 wt.% NaCl aqueous solution, and with a 20-rpm impeller speed:

• The electrolyte’s CaCO3 saturation level, [Ca2+], bulk pH, and temperature are crucial parameters in studying the effect of Ca2+ on CO2 corrosion. Ignoring the influence of such parameter(s) is one of the main reasons for the existing discrepancies in the available literature on this topic.

• Fe3C played a critical role in precipitation of FexCayCO3, x+y=1, and FeCO3. The mole fraction of Ca in FexCayCO3 depended on [Ca2+] in the bulk and the experimental conditions.

• The protective behavior of the surface layers was mainly due to the formation of FeCO3 adjacent to the steel surface and not FexCayCO3 nor CaCO3 scale. However, the presence of FexCayCO3 and/or CaCO3 scale facilitated FeCO3 precipitation and enhanced its protectiveness by acting as a mass transfer barrier of Fe2+ outwards from the steel and H+ towards the steel.

• Although CaCO3 is isostructural with FeCO3, CaCO3 scale was not protective against further corrosion where FeCO3 was protective. Ca2+ comes from bulk solution while Fe2+ comes from the corroding steel surface. Therefore, FeCO3 has superior adherence to steel. This makes FeCO3 protective while CaCO3 is not.

• No localized corrosion was observed in the presence of different concentrations of Ca2+ and uniformly formed CaCO3 scale in the conducted experimental conditions. However, ununiformed precipitation of CaCO3 scale along with change of water chemistry on the steel surface might lead to localized corrosion 14.

2


3

Now that a better understanding of the influence of Ca2+ on the CO2 corrosion of mild steel has been achieved, it is relevant to first explore the individual effect of Mg2+ and then the effect of the simultaneous presence of Ca2+ and Mg2+ on corrosion of mild steel to simulate the oilfield brines. In the present study, the results related to two experimental scenarios are presented. In the first scenario, the electrolyte is saturated with respect to MgCO3 with low and high [Mg2+]. In the second scenario, the electrolyte is primarily saturated with respect to CaCO3 and Mg2+ are introduced at a mass ratio Ca2+ over Mg2+ of 7 (a typical mass ratio observed in oilfields9). Precipitation kinetics of CaCO3 is relatively fast and CaCO3 was successfully precipitated on the substrate surface, as shown experimentally in a previous publication by the authors.12 However, precipitation kinetics of MgCO3 are very low and MgCO3 did not precipitate in the laboratory environments, even from highly supersaturated solutions with respect to MgCO3, as is described later on in this publication (cathodic polarization of the working electrode did not help, either). Therefore, unlike for the CaCO3 scale, it proved impossible to precipitate and study the effect of solid MgCO3 scale on CO2 corrosion in the current experimental conditions. However, as mentioned earlier, a pure MgCO3 scale is also rarely seen in oilfields.

The main objectives of this work are to:

• Investigate the effect of low and high concentrations of Mg2+ on the CO2 corrosion

mechanism of mild steel while the electrolyte is saturated with respect to MgCO3 at

applicable pH conditions (Scenario 1).

• Investigate the effect of the simultaneous presence of Mg2+ and Ca2+ (with Ca2+/Mg2+ = 7,

mass ratio) on the CO2 corrosion mechanism of mild steel in a solution saturated with

respect to CaCO3 (Scenario 2).

Special efforts were made to control the water chemistry and flow characteristics in the experimental glass cell system, ensuring reproducibility and reliability of the data.

EXPERIMENTAL SETUP, TEST MATRICES, AND METHODOLOGY

Experimental Setup The experimental setup is able to control [Fe2+] and [H+] by means of two independent ion-exchange resin loops (Na-form and H-form, respectively).11 It should be mentioned that for experiments with lower pH value (5.5) and higher Mg2+ concentration (4,200 ppm), [Fe2+] was not controlled due to the fact that the Na-form resin is unable to differentiate between Mg2+ and Fe2+ and takes up both dipositive cations. Therefore, after a short period, the resin becomes saturated with Mg2+ and Fe2+ and loses its effectiveness. Despite the inability to control [Fe2+] at pH 5.5, the FeCO3 saturation levels at pH 6.2 and 5.5 were still similar (due to the higher solubility of FeCO3 at pH 5.5 as compared to pH 6.2). Therefore, the comparison of the results for both pH value is still valid. More detailed information about the Na-form and H-form ion-exchange resins are presented in a previous publication by the authors11 and also in a paper by Zhonh, et al.15 The experimental setup is equipped with an impeller to create uniform mass transfer across the specimens. It also includes stationary specimen holders, instead of the more conventional rotating cylinder electrode and hanging weight loss specimens, installed around a central impeller. Moreover, the design of specimen holders eliminate the risk of oxygen contamination during specimen retrieval from the test solution The flow characteristics of this experimental setup have been reported in a recent publication by the authors.11

3


4

Test Matrices To address the first objective described earlier, MgCO3-saturated experiments were conducted and repeated for two different bulk solution pH values, 6.2 and 5.5. Using different solution pH values permitted to use different dissolved Mg2+ concentrations while the solution was saturated with respect to MgCO3. For each Mg2+-containing experiment, a baseline experiment was performed, in an otherwise identical water chemistry condition. An MgCl2 salt (ACROS Organics™) was used to prepare Mg2+-containing solutions.

Table 1

Test matrix for experiments saturated with MgCO3 at pH 6.2 and pH 5.5

Parameter Description

Material & Specimen Type

UNS G10180 with ferritic-pearlitic microstructure flat square specimen with 1.5 cm2 exposed surface area

Temperature 80C

pCO2 0.53 bar

pH 6.2 5.5

Electrolyte

Baseline

Investigating effect of Mg2+

(low concentration)

Baseline

Investigating effect of Mg2+

(high concentration)

1 wt.% NaCl +NaHCO3

(Ionic Strength ≈0.18 M)

1 wt.% NaCl +MgCl2


1 wt.% NaCl +NaHCO3+

NaClO4 (Ionic Strength ≈

0.6 M)

1 wt.% NaCl +MgCl2


[Mg2+] 0 100 ppm 0 4,200 ppm

𝐒𝐌𝐠𝐂𝐎𝟑 0 1 0 1

𝐒𝐅𝐞𝐂𝐎𝟑 < 10

Dissolved O2 <5 ppb

Reference Electrode

Saturated Ag/AgCl

Impeller Rotation Speed

20 rpm

Mass Transfer Conditions

Equivalent to 0.5 m/s in a 0.1m ID pipe

Electrochemical Techniques

LPR, OCP, EIS

Surface Analysis

Techniques XRD, SEM/EDS

Test Duration 7 days

The experimental conditions presented in Table 2 was employed to conduct experiments related to the second scenario when Mg2+ and Ca2+ are simultaneously present in the electrolytes. The solution was initially saturated with respect to CaCO3 and undersaturated with respect to MgCO3. An excess amount of granular CaCO3, solid powder, was present at the beginning of the experiments to ensure that the solution remained saturated with respect to CaCO3. Mg2+

was then added to the system using the MgCl2 salt, targeting a mass ratio of Ca2+/Mg2+ =7. In

4


5

the conducted experimental conditions, the [Ca2+] was ~6000 ppm and the [Mg2+] was ~857 ppm. Previous publications by the authors11,12 provide more information about the preparation of the CaCO3-saturated solution with excess CaCO3 powder.

Table 2

Experimental condition to study the effect of simultaneous presence of [Mg2+] and [Ca2+] on CO2 corrosion at pH 5.5

Parameter Description

Material & Specimen Type

UNS G10180 with ferritic-pearlitic microstructure flat square specimen with 1.5 cm2 exposed surface area

Temperature 80C

2pCO 0.53 bar

pH 5.5

Electrolyte

Baseline Ca2+/Mg2+=7 (mass ratio)

1 wt.% NaCl + NaHCO3 + NaClO4

(Ionic Strength≈0.6 M)

0.85 wt.% NaCl + CaCO3 + MgCl2

(Ionic Strength≈0.6 M)

]2+[Mg 0 857 ppm, 𝑆𝑀𝑔𝐶𝑂3≈0.2

[Ca2+] 0 6,000 ppm, 𝑆𝐶𝑎𝐶𝑂3≈1

𝐒𝐅𝐞𝐂𝐎𝟑 < 10

2Dissolved O < 5 ppb

Reference Electrode Saturated Ag/AgCl

Impeller Rotation Speed 20 rpm

Mass Transfer Conditions

Equivalent to 0.5 m/s in a 0.1m ID pipe

Electrochemical Techniques

LPR, OCP, EIS

Surface Analysis Techniques

XRD, SEM/EDS

Experiment Duration 7 days

Methodology For each experiment, regardless of the scenarios, [Fe2+] was measured twice daily by spectrophotometry using phenanthroline as the complexation reagent. The impeller rotational speed was set to 20 rpm, conferring a mass transfer rate similar to that encountered in a 0.1m ID pipe with a flow velocity of 0.5 m/s.11 The specimens were made of UNS G10180 mild steel that has a ferritic-pearlitic microstructure. The electrochemical and weight loss specimens were wet-polished with silicon carbide abrasive papers up to 600 grit. Following the polishing process, the specimens were rinsed with isopropanol and placed in an ultrasonic cleaner for 2 minutes to remove any possible debris from the steel surface. Finally, they were dried by cold air and weighed prior to immersion into the test solutions. A three-electrode system, including working, counter, and reference electrodes, along with a Gamry Reference600™ potentiostat, was used

5


6

to conduct electrochemical measurements. The corrosion rate was measured at least twice daily using the linear polarization resistance (LPR) method. The open circuit potential (OCP) was also recorded. The duration of each experiment was 7 days. Two specimens were retrieved from the glass cell at days 2, 4, and 7 from each experiment to obtain the weight loss (WL) as well as to conduct surface characterization using techniques including scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD). The polarization resistance measurements obtained by LPR included determination of solution resistance. Electrochemical impedance spectroscopy (EIS) was used to measure the solution resistance and the polarization resistance at the metal/solution interface was consequently compensated. In this study, a B value of 26 mV was used in the Stern-Geary equation to convert the experimental polarization resistance to a corrosion rate. This value is commonly accepted in CO2 environments, but is not based on any specific Tafel slopes since the corrosion mechanism is not strictly charge transfer controlled. Instead, this B value was determined by best fit comparison between current densities and weight loss measurements.16,17

RESULTS AND DISCUSSION

In this section, the experimental results of the two scenarios are presented. First, results of experiments saturated with MgCO3 conditions at two different pHs are discussed. Second, experimental results of the second scenario (simultaneous presence of Ca2+and Mg2+) are presented.

Scenario 1: Effect of MgCO3-Saturated Solution at pH 6.2 & 5.5

Results and Discussion Related to the Experiments Conducted at pH 6.2

Figure 1 presents the comparison of corrosion rate (CR) by the LPR for MgCO3-saturated experiments (Mg2+ ca. 100 ppm) and the baseline experiment, at pH 6.2. The error bars on the graphs provided in Figure 1, and other graphs throughout this paper, represent maximum and minimum values at each average point from two separate experiments. For each experiment, specimens at day 2, 4, and 7 were retrieved from the test solution to conduct weight loss and surface characterization (such specimens are indicated with gray squares in Figure 1). Based on Figure 1, it seems that the presence of Mg2+ accelerated the CR in the active period compared to the baseline. However, after 4 days of exposure, the corrosion rate became similar for experiments with and without Mg2+. At this time (pseudo-passivation), the corrosion product layers have fully developed and, as a result, the iron dissolution reaction (anodic reaction) is significantly retarded compared to the initial stage of the corrosion process (active corrosion period). The three corrosion periods indicated in Figure 1 are thoroughly explained in a previous publication.11 In conclusion, the presence of Mg2+ at ca. 100 ppm did not alter the protectiveness of corrosion products when they are fully developed in the pseudo-passivation period. The only difference was the higher initial CR experienced in the solution saturated with MgCO3 in the active corrosion period.

6


7

Figure 1: LPR corrosion rate over time with and without Mg2+ at pH 6.2

Specimens were retrieved from the test solutions on day 2, 4, and 7 of the experiments for surface layer characterization and measuring CR by the WL technique. Figure 2 shows a comparison of time-average CR by WL (bar chart) and LPR (line chart) at different exposure times for experiments with and without Mg2+. Both LPR and WL methods show that CR decreases over time probably due to formation and development of corrosion product layers. Another observation is that although the initial CR in the presence of Mg2+ seems slightly higher than the baseline experiment (see CR by LPR in Figure 1), the cumulative CRs for both experiments are similar at the first two days; based on Figure 2. The thickness of the Fe3C layer is also similar in both cases (see Figure 3). Fe3C has been reported to act as an anchoring site for nucleation and crystal growth, collectively termed precipitation, of FeCO3.18,11

Figure 2: Cumulative corrosion rate by WL (bar chart) and LPR (line chart) techniques over time

for solution with and without Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 6.2, ionic strength 0.18 M, 1 wt/% NaCl)

7


8

Figure 3: Evolution of Fe3C over time with and without presence of Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 6.2, ionic strength 0.18 M, 1 wt.% NaCl)

A comparison of open circuit potential (OCP) for experiments with and without Mg2+ is provided in Figure 4. The OCP trend with and without Mg2+ are identical with less negative OCP achieved at the end of both experiments (pseudo-passivation period) due to a better protectiveness offered by the corrosion products.

Figure 4: Comparison of OCP over time for UNS G10180 exposed to solutions with and without

Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 6.2, ionic strength 0.18 M, 1 wt.% NaCl)

Surface layers were characterized by SEM and the chemical compositional of the corrosion products and scales were analyzed by EDS. Furthermore, XRD was used to characterize the corrosion products on the steel surface. SEM cross-sectional and top view images of surface layers developed without and with Mg2+ are provided in Figure 5 and Figure 6, respectively. The yellow arrows on the cross-sectioned specimens indicate the calculated metal loss depth based on WL corrosion rate. Such values were typically slightly greater than the measured corrosion product thickness. Overall, the morphology of surface layers is similar with and without the presence of Mg2+. One difference is that the Fe3C thickness in the presence of Mg2+ is higher in the first 2 days because of the higher corrosion rate. A greater thickness of Fe3C means more favorable water chemistry for

8


9

precipitation of FeCO3 within its pores. Upon precipitation of FeCO3, CR would drop; Figure 1 is indicative of this phenomenon. A complete explanation for the role of Fe3C in CO2 corrosion over time is provided in a previous publication.11,12

Figure 5: SEM images (top and cross-section view) of the development of surface layers over

time for the experiment without presence Mg2+ (𝑆𝑀𝑔𝐶𝑂3 = 0) at 80C, pCO2 0.53 bar, 0.5 m/s, pH

6.2, ionic strength 0.18 M, 1 wt.% NaCl


time for the experiment with presence of 100 ppm Mg2+ (𝑆𝑀𝑔𝐶𝑂3 = 1) at 80C, pCO2 0.53 bar, 0.5

m/s, pH 6.2, ionic strength 0.18 M, 1 wt.% NaCl

Figure 7 depicts a comparison of cross-sectional EDS line scan analysis of the surface layers developed on specimens exposed to solutions with and without Mg2+ after 7 days. FeCO3 is the only corrosion product in the absence of Mg2+ (Figure 7 (a)). Even in the presence of Mg2+, FeCO3 was the only corrosion product existing adjacent to the steel surface. This inner layer was responsible for the final decrease in the corrosion rate (Figure 7 (b)). The formation of a

9


10

mixed iron-magnesium carbonate, as an outer layer, was confirmed by these analyses in the presence of Mg2+. However, the mole fraction of Fe is significantly dominant over Mg in such a mixed metal carbonate.

Figure 7: EDS line scan analysis of the surface layers developed on the specimens after 7 days

of exposure to electrolytes without (a) and with (b) Mg2+ at 80C, pCO2 0.53 bar, 0.5 m/s, pH 6.2, ionic strength 0.18 M, 1 wt.% NaCl

Figure 8 shows the XRD data obtained for the specimens recovered from a MgCO3-saturated solution with a Mg2+ concentration of ca. 100 ppm after 2 days (black line), 4 days (yellow line), and after 7 days (purple line) of exposure. XRD data shows that after 2 days of exposure, α-Fe and Fe3C peaks are dominant over weak peaks related to FeCO3. However, after 4 and 7 days of exposure, the intensity of peaks related to α-Fe and Fe3C are noticeably diminished and, instead, peaks associated with FeCO3 are intensified. The more intense peaks for FeCO3

associated with (104) and (116) Miller planes located at 32.07 and 52.7 degrees 2 (CuKα radiation) are more easily recognizable after 4 and 7 days of exposure. However, FeCO3 peaks are now broadened and shifted slightly towards the MgCO3 reference peaks (green lines) after longer exposure time. Nevertheless, the incorporation of Mg2+ into the lattice of FeCO3 was very limited compared to the similar condition with the presence of Ca2+.11

10


11

Figure 8: XRD patterns of surface layers detected on the steel surface after 2, 4, and 7 days of

exposure to an electrolyte saturated with MgCO3 at 80C, pCO2 0.53 bar, 0.5 m/s, pH 6.2, ionic strength 0.18 M, 1 wt.% NaCl

The susceptibility of the specimens to localized corrosion when exposed to solution with and without Mg2+ was also evaluated. Profilometry of the specimen surfaces was performed after removing corrosion product layers by Clarke solution 19. No localized corrosion was observed for experiments conducted at bulk pH 6.2 with ca. 100 ppm Mg2+. Figure 9 illustrates an example of profilometry of a specimen exposed for 7 days to a solution saturated with respect to MgCO3.

Figure 9: Example of profilometry of the specimen after removing the corrosion products (7 days

of exposure to solution saturated with MgCO3 at pH 6.2, 80C, pCO2 0.53 bar, ionic strength 0.18 M, 1 wt.% NaCl, Mg2+ ~ 100 ppm

11


12

Results and Discussion Related to the Experiments Conducted at pH 5.5

An additional experiment was conducted in similar conditions but at a higher concentration of Mg2+, while the solution was still saturated with MgCO3, to better resemble oilfield conditions. The experiment was conducted with 4,200 ppm Mg2+ at pH 5.5 (Table 2). Figure 10 shows the trend of LPR corrosion rate over time with and without Mg2+. The three corrosion periods (active corrosion, nucleation/growth, pseudo-passivation) are identified similarly to the experiments conducted at pH 6.2. The final corrosion rates are higher compared to pH 6.2 because of the more aggressive solution (lower pH). Figure 11 shows cumulative corrosion rate with and without Mg2+. According to WL results, the presence of 4,200 ppm Mg2+ leads to a higher average CR, especially after the first two days of exposure. This is reflected in the higher Fe3C thickness as shown in Figure 12. The difference in CR becomes less pronounced as exposure time increases. The OCP variation over time for experiments conducted at pH 5.5 is shown in Figure 13.

Figure 10: Trend of corrosion rate obtained by LPR over time with and without presence of Mg2+

at pH 5.5

12


13

Figure 11: Cumulative corrosion rate by WL (bar chart) and LPR (line chart) techniques over

time for solution with and without Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6 M, 1 wt.% NaCl)

Figure 12: Evolution of Fe3C over time with and without presence of Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6 M, 1 wt.% NaCl)

Figure 13: Comparison of OCP over time for UNS G10180 exposed to solutions with and

without Mg2+ (80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.68 M, 1 wt.% NaCl)

13


14

SEM cross-sectional and top view images of surface layers developed without and with the presence of Mg2+ are shown in Figure 14 and Figure 15, respectively. The morphology and composition of corrosion products were similar to those obtained at pH 6.2 with lower [Mg2+]. A greater thickness of corrosion products was achieved for experiments at pH 5.5 due to the higher corrosion rate.


time for the experiment without presence of Mg2+ (𝑆𝑀𝑔𝐶𝑂3 = 0) at 80C, pCO2 0.53 bar, 0.5 m/s,

pH 5.5, ionic strength 0.6 M, 1 wt.% NaCl


time for the experiment with presence of 4,200 ppm Mg2+ (𝑆𝑀𝑔𝐶𝑂3 = 1) at 80C, pCO2 0.53 bar,

0.5 m/s, pH 5.5, ionic strength 0.6M, 1 wt.% NaCl

14


15

Figure 16 shows a comparison of cross-sectional EDS line scan analysis of the surface layers developed on specimens exposed to solutions with and without 4,200 ppm Mg2+ after 7 days. FeCO3 is the only corrosion product in the absence of Mg2+ (see Figure 16 (a)). Figure 7 (b), considering 4,200 ppm of Mg2+, shows no significant Mg2+ incorporation into the lattice of FeCO3. Likewise, for the experiment with lower [Mg2+], a solid solution of FexMgyCO3 formed on the steel surface with x>>y. Formation of a bi-layer of corrosion product on the specimen is confirmed from Figure 7 (b).

Figure 16: EDS line scan analysis of the surface layers developed on the specimens after 7

days of exposure to electrolytes without (a) and with (b) 4,200 ppm Mg2+ at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M, 1 wt.% NaCl

Unlike the experiments conducted at low [Mg2+], localized corrosion was observed with high [Mg2+] at the lower pH of 5.5. Corrosion products were occasionally ruptured on the steel surface after 4 and 7 days of exposure (Figure 20). Beneath such damaged layers, localized corrosion was observed after removal of corrosion products. A thick and fragile layer of Fe3C, in comparison to the baseline experiment (see Figure 12), was more vulnerable to rupture before FeCO3 could precipitate within its pores and confer more mechanical support. Figure 17 shows a top and cross-section view of such damage in corrosion products possibly caused by fluid flow.

15


16

Top View Cross-section view

Figure 17: SEM images (top and cross section view) of ruptured corrosion products observed on

specimens exposed to MgCO3-satuarted solution with 4,200 ppm Mg2+ at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M, 1 wt.% NaCl

Figure 18 shows profilometry of specimens after removal of corrosion products by Clarke solution.19 Localized corrosion was observed at locations where the corrosion product layer was damaged. The number of localized attack sites was limited, and the fractures were relatively wide. That is why the pitting ratio (maximum penetration rate over uniform corrosion rate) of such attacks is small. This observation of localized corrosion was repeatable in the presence of high concentration of Mg2+. Table 3 provides more information regarding maximum pit depth and pitting ratio for two different experimental conditions where specimens where exposed to MgCO3-saturated solution with 4,200 ppm Mg2+. The maximum pitting ratio being lower than 2, this corrosion attack does not qualify as “pitting corrosion”.

2-day exposure 4-day exposure 7-day exposure

Figure 18: Profilometry of specimens after removing of corrosion product layers (MgCO3-

satuarted solution with 4,200 ppm Mg2+ at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M, 1wt% NaCl)

16


17

Table 3

Evolution of localized attacks on specimens exposed to 4,200 ppm Mg2+ at pH 5.5 R

ep

eata

bilit

y o

f lo

ca

lize

d c

orr

osio

n

in p

res

en

ce o

f h

igh

[M

g2+]

at

pH

5.5

Experiment #A

Exposure time (day)

Max pit (µm)

Max penetration rate (mm/y)

Pitting ratio

2 8 1.46 0.12

4 140 12.78 1.08

7 237 12.36 1.49

Experiment #B (repeat of A)

Exposure time day Max pit (µm)

Max penetration rate (mm/y)

Pitting ratio

2 12 2.19 0.17

4 200 16.22 1.59

7 216 11.26 1.56

Based on the experimental results, a mechanism for the localized corrosion of mild steel exposed to high concentration of Mg2+ is proposed as illustrated in Figure 19:

1) 1018 carbon steel is exposed to solution saturated with MgCO3 and CO2 with bulk pH

5.5.

2) Fe oxidatively dissolves and Fe2+ is released into solution. Consequently, a porous Fe3C

network is left behind on the steel surface and is growing, as a residue, over time. It is

postulated that the presence of Mg2+ has hindered nucleation/growth of FeCO3 crystals

within Fe3C pores. Therefore, corrosion rate is higher compared to Ca2+-containing

experiments at this stage of the corrosion process. Other researchers have also reported

the inhibition effect of Mg2+ on carbonate nucleation in different aqueous systems.3,20

3) The fragile Fe3C layer reach a critical thickness. Occasional breakage occurs possibly

due to internal stress within the Fe3C network as well as to flow effects.

4) Continuation of corrosion and Fe3C development.

5) Achievement of favorable water chemistry for nucleation/precipitation of FeCO3 within

Fe3C pores. At this stage, the very favorable water chemistry for precipitation of FeCO3

overcomes the inhibiting effect of Mg2+ on FeCO3 precipitation and as a result the

general corrosion rate decreased.

6) Occurrence of localized corrosion where the steel is not covered with FeCO3 (locations

with damaged Fe3C).

7) Retardation of localized corrosion rate due to further precipitation of FeCO3.

8) Further development of FeCO3 (inner layer) and FexMgyCO3, x>>y (outer layer)

17


18

(1)

CS

10

18 e

xp

ose

d t

o

4,2

00

pp

m M

g2+

(2)

inh

ibitio

n e

ffect

of

Mg

2+

on F

eC

O3, d

eve

lop

me

nt o

f

a th

ick a

nd

fri

gile

Fe

3C

(3)

oca

sio

na

l b

rea

ka

ge

of

Fe

3C

(4)

co

ntin

uatio

n o

f

co

rro

sio

n a

nd F

e3C

exp

osu

re

(5)

favo

rab

le w

ate

r

ch

em

istr

y w

ith

in F

e3C

netw

ork

, pre

cip

ita

tio

n o

f

Fe

CO

3

(6)

unco

ve

red

are

a,

loca

lize

d c

orr

osio

n

(7)

pre

cip

ita

tio

n o

f F

eC

O3

with

in p

it,

pit d

ea

th

(8)

deve

lopm

ent

of

Fe

CO

3

(re

d c

olo

r) a

nd m

ixe

d

Fe

xM

gyC

O3 (

ye

llow

co

lor)

with

in F

e3

C p

ore

s.

Figure 19: Proposed mechanism for the localized corrosion observed in presence of high [Mg2+]

for mild steel at 4,200 ppm Mg2+, 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M, 1 wt.% NaCl

Scenario 2: Effect of the Simultaneous Presence of Mg2+ and Ca2+ on CO2 Corrosion at pH 5.5 (Test Condition Presented in Table 2) In the previous section, the effect of Mg2+ and Ca2+ was studied separately. However, Mg2+ and Ca2+ are simultaneously observed in oilfield brines and it is consequently relevant to study the effect of these ions concurrently. In the current experimental scenario, the solution was

Mild steel-10180

ferritic-pearlitic structure

18


19

saturated with CaCO3 (Ca2+ concentration ca. 6000 ppm) with addition of ca. 857 ppm Mg2+ as MgCl2 salt (rendering a mass ratio of 7, Ca2+/Mg2+) based on conditions described in Table 2. The trend of LPR corrosion rate over time with different water chemistry at pH 5.5 is presented in Figure 20. The presence of Mg2+ in a solution saturated with CaCO3 (yellow line) seems to increase the general corrosion rate (compared to Ca2+-containing solution without Mg2+, dark blue line). Cumulative corrosion rate at day 2, 4 and 7 exposure times to different solutions obtained by WL method is provided in Figure 20. WL results, which agree with the LPR data, show that corrosion in the presence of 4,200 ppm Mg2+ is higher than the other scenarios at each measuring point. The addition of 857 ppm Mg2+ to a CaCO3-saturated solution (yellow line) clearly increases the corrosion rate compared to a CaCO3-saturated solution with no Mg2+ (dark blue line).

Figure 20: Trend of corrosion rate obtained by LPR over time at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M [baseline (red), 6000 ppm Ca2+ (dark blue), Ca2+/Mg2+=7 (yellow),

and 4200 ppm Mg2+ (light blue)]

Figure 21: Cumulative corrosion rate obtained by WL over time at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M [baseline (red), 6000 ppm Ca2+ (dark blue), Ca2+/Mg2+=7 (yellow),

and 4200 ppm Mg2+ (light blue)]

19


20

SEM and EDS analysis revealed that a lesser quantity of carbonates precipitated on top of the Fe3C layers with the simultaneous presence of Mg2+ and Ca2+, compared to a solution solely saturated with CaCO3. It seems that the presence of Mg2+ diminished precipitation of carbonate crystals. Figure 22 illustrates SEM/EDS data, for comparison, for two solutions: (a) Ca2+/Mg2+=7, SCaCO3=1 and (b) Ca2+= 6000 ppm, SCaCO3=1.

Figure 22: Comparison of SEM/EDS results for two solutions at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.68 M [(a); Ca2+/Mg2+=7, SCaCO3=1 & (b) Ca2+= 6000 ppm, SCaCO3=1(ref. 12)]

Figure 23 shows EDS line scan analysis of a cross-sectioned specimen after 7 days of exposure to the solution saturated with CaCO3 and in the presence of 857 ppm Mg2+. This analysis revealed that a mixed solid solution of FexCayMgzCO3 (x>y, z is trace) is formed at the outer layer whereas the inner layer was a pure FeCO3 phase.

20


21

Figure 23: EDS line scan analysis of the surface layers developed on a specimen after 7 days of

exposure to a CaCO3-saturated electrolyte with 857 ppm Mg2+ at 80C, pCO2 0.53 bar, 0.5 m/s, pH 5.5, ionic strength 0.6M

CONCLUSIONS

The following conclusions can be drawn: o Localized corrosion was observed in the presence of a high concentration of Mg2+. A

descriptive model is provided for this finding.

o The presence of low and high concentrations of Ca2+ (at pH 6.2 and 5.5, respectively),

as far as solution is under saturated or saturated with respect to CaCO3, would not

endanger protectiveness of corrosion products and no localized corrosion was observed.

11, 13

o A uniformly formed CaCO3 scale on the steel surface is not protective, however, its

presence facilitates the formation of protective FeCO3 as the inner-layer corrosion

product. Partial precipitation of CaCO3 on the substrate might lead to localized

corrosion.12

o No localized corrosion was observed for electrolytes with a simultaneous presence of

Ca2+ and Mg2+.

o The presence of Mg2+ adversely influenced precipitation of carbonates (FeCO3 and

CaCO3) at the outer layer in solutions saturated with respect to CaCO3.

ACKNOWLEDGMENTS

The author would like to thank the following companies for their financial support: Anadarko,

Baker Hughes, BP, Chevron, Clariant Corporation, CNOOC, ConocoPhillips, DNV GL,

ExxonMobil, M-I SWACO (Schlumberger), Multi-Chem (Halliburton), Occidental Oil Company,

PTT, Saudi Aramco, SINOPEC (China Petroleum) and TOTAL.

REFERENCES

[1] Z. Dai, A. Kan, F. Zhang, and M. Tomson, “A Thermodynamic Model for the Solubility Prediction of Barite, Calcite, Gypsum, and Anhydrite, and the Association Constant

21


22

Estimation of CaSO4 Ion Pair up to 250 °C and 22000 psi,” J. Chem. Eng. Data, vol. 60, no. 3, pp. 766–774, 2015.

[2] D. H. Case, F. Wang, and D. E. Giammar, “Precipitation of Magnesium Carbonates as a Function of Temperature, Solution Composition, and Presence of a Silicate Mineral Substrate,” Environ. Eng. Sci., vol. 28, no. 12, pp. 881–889, 2011.

[3] M. Hänchen, V. Prigiobbe, R. Baciocchi, and M. Mazzotti, “Precipitation in the Mg-carbonate system—effects of temperature and CO2 pressure,” Chem. Eng. Sci., vol. 63, no. 4, pp. 1012–1028, 2008.

[4] Z. D. Cui, S. L. Wu, S. L. Zhu, and X. J. Yang, “Study on corrosion properties of pipelines in simulated produced water saturated with supercritical CO2,” Appl. Surf. Sci., vol. 252, no. 6, pp. 2368–2374, 2006.

[5] C. de Waard, U. Lotz, and D. E. Milliams, “Predictive Model for CO2 Corrosion Engineering in Wet Natural Gas Pipelines,” Corrosion, vol. 47, no. 12, pp. 976–985, 1991.

[6] K. Gao, F. Yu, X. Pang, and G. Zhang, “Mechanical Properties of CO2 Corrosion Product Scales and Their Relationship to Corrosion Rates,” Corros. Sci., vol. 50, no. 10, pp. 2796–2803, 2008.

[7] B. R. Linter and G. T. Burstein, “Reactions of pipeline steels in carbon dioxide solutions,” Corros. Sci., vol. 41, no. 1, pp. 117–139, 1999.

[8] J. Han, J. W. Carey, and J. Zhang, “Effect of sodium chloride on corrosion of mild steel in CO2-saturated brines,” J. Appl. Electrochem., vol. 41, no. 6, pp. 741–749, 2011.

[9] P. E. Dresel and A. W. Rose, “Chemistry and Origin of Oil and Gas Well Brines in Western Pennsylvania,” The Pennsylvania State University, Open-File Report OFOG 10–01.0, 2010.

[10] H. Mansoori, D. Young, B. Brown, and M. Singer, “Influence of Calcium and Magnesium Ions on CO2 Corrosion of Carbon Steel in Oil and Gas Production Systems-A Review,” J. Nat. Gas Sci. Eng., vol. 59, pp. 287–296, 2018.

[11] H. Mansoori, D. Young, B. Brown, S. Nesic, and M. Singer, “Effect of CaCO3-Saturated Solution on CO2 Corrosion of Mild Steel Explored in a System with Controlled Water Chemistry and Well-Defined Mass Transfer Conditions,” Corros. Sci., vol. 158, pp. 108078

2019. [12] H. Mansoori, D. Young, B. Brown, S. Nesic, and M. Singer, “Effect of Calcium Ions and

CaCO3 Scale on the CO2 Corrosion Mechanism of Mild Steel,” in NACE International, Paper No. 13000, 2019.

[13] H. Mansoori, D. Young, B. Brown, S. Nesic, and M. Singer, “CO2 Corrosion of Mild Steel Exposed to CaCO3-Saturated Aqueous Solutions,” Corrosion, vol. 75, no. 11, pp. 1281-1284, 2019.

[14] S. N. Esmaeely, D. Young, B. Brown, and S. Nesic, “Effect of Incorporation of Calcium into Iron Carbonate Protective Layers in CO2 Corrosion of Mild Steel,” Corrosion, vol. 73, no. 3, pp. 238–246, 2016.

[15] X. Zhong, B. Brown, W. Li, S. Nesic, and M. Singer, “How to Maintain a Stable Solution Chemistry when Simulating CO2 Corrosion in a Small Volume Laboratory System,” in NACE International, Paper No. 7780, 2016.

[16] Y. Yang, B. Brown, M. E. Gennaro, and B. Molinas, “Mechanical Strength and Removal of a Protective Iron Carbonate Layer Formed on Mild Steel in CO2 Corrosion,” in NACE International, Paper No. 10383, 2010.

[17] F. Madani Sani, B. Brown, Z. Belarbi, and S. Nesic, “An Experimental Investigation on the Effect of Salt Concentration on Uniform CO2 Corrosion,” in NACE International, Paper No.13026, 2019.

[18] F. Farelas, M. Galicia, B. Brown, S. Nesic, and H. Castaneda, “Evolution of dissolution processes at the interface of carbon steel corroding in a CO2 environment studied by EIS,” Corros. Sci., vol. 52, no. 2, pp. 509–517, 2010.

22


23

[19] ASTM G1, “Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test.” ASTM International, 2011.

[20] S.H. Lin and S. C. Dexter, “Effects of Temperature and Magnesium Ions on Calcareous Deposition,” Corrosion, vol. 44, no. 9, pp. 615–622,1988.

23


Effect of CaCO3 Aqueous Saturation on CO2 Corrosion of ...

Documents

Transcript of Effect of CaCO3 Aqueous Saturation on CO2 Corrosion of ...